[Mmpie]

That is exactly what I had in mind zee, albeit I was thinking about trying to make my own slightly smaller version.

Trying to get my X and Y square today, it was absolutely miles out. By drilling out the gantry sides mounting holes to 9mm (8mm for the ones on the side which also hold the cable chain) I got it to 1.14mm across the 355mm span of the X, no amount of forcing it into position will change that. I think the gantry side arms might just be twisting now and the ball screw/rails are holding firm in their position? Any ideas anyone?

Can't believe I never noticed how far it was out before. Measured some bigger panels I cut early days and they are indeed parallelograms rather than squares and rectangles.

Must have been a new guy made my machine. It really is a piece of ****.


Sent from my iPhone using Tapatalk

[zeeflyboy]

You need to take it apart and get the beam that spans the carriages square first... I would suggest using a machinist square to check if the front piece is square to the Y-rails, then if that looks good you can use the front plate as your square reference - loosen up the carriage screws and clamp the cross beam to the front plate. Re-tighten the carriage screws (also worth doing the ball nut btw) and then release the clamp.

You can then do the finer adjustment like I showed in a previous post by tapping the gantry with the bolts at the base loosened. Once you've squared up that bottom beam it should be relatively small adjustments required for the gantry - I didn't need to enlarge any holes.

[Mmpie]

Not what I wanted to hear... hoped to avoid doing that, it took ages to get the bed shimmed

Is there plenty of wiggle room in those holes in the lower beam? Or a case of opening them up? I've had it off once before but I can't recall. I do remember it having milled pockets underneath for the carriages and ball nut to sit in. Not sure they will have room to make up over 1mm difference, from memory they were quite a snug fit.


Sent from my iPhone using Tapatalk

[zeeflyboy]

Pockets are over sized iirc, I had no issues squaring it up without messing around opening up bolt holes etc.

Biggest issue in squaring it for me was how bendy the gantry is.

[Mmpie]

You need to take it apart and get the beam that spans the carriages square first... I would suggest using a machinist square to check if the front piece is square to the Y-rails, then if that looks good you can use the front plate as your square reference - loosen up the carriage screws and clamp the cross beam to the front plate. Re-tighten the carriage screws (also worth doing the ball nut btw) and then release the clamp.

You can then do the finer adjustment like I showed in a previous post by tapping the gantry with the bolts at the base loosened. Once you've squared up that bottom beam it should be relatively small adjustments required for the gantry - I didn't need to enlarge any holes.

I did as you suggested today, thankfully the y rails are both perfectly square to the front and back plates. Used a bit of tooling plate as a flat surface to space the gap between the back plate and the beam, clamped and rebolted (12nm on every carriage bolt with the torque wrench) easily enough. The beam was skewed considerably over the 300mm length of the tooling plate to start with, around 1.5mm. Results at first looked perfect but after half hour or so it had sprung back out by 0.28mm. Leaving it til morning to see how far it goes. I'm thinking tomorrow to shim the side which is closer to the back plate by the distance of whatever the gap is and repeat? Or possibly even further, as if it does go beyond there's plenty of movement on the side arm connections to correct. Is this what you'd do?

As I mentioned yesterday I think I've reached the max movement at the gantry sides in the current direction it's out to correct it there.



Sent from my iPhone using Tapatalk

[zeeflyboy]

I must admit, I don't quite get how it springs back!

Do you perhaps need to enlarge the holes slightly? Did you re-seat the ball screw nut as well as the carriages?

[Mmpie]

I must admit, I don't quite get how it springs back!

Do you perhaps need to enlarge the holes slightly? Did you re-seat the ball screw nut as well as the carriages?

Yes I reseated the lot. Probably could do with opening the holes up a bit, most the bolts still dropped straight through and turned easily, but one or two were rubbing a bit. Although that was true for both sides of the beam and one end stayed put (or at least to within 0.04mm), it was only the side that I was trying to force into place that sprang back.


Sent from my iPhone using Tapatalk

[Mmpie]

I managed to get it just about square today. Got the beam sorted, I couldn't get the drill in to the bolts at the edge without removing the gantry, which meant removing the spindle and cable chain (again). Not difficult but I opted to try the lazy route first by shimming out one side and it worked a treat, the other side sprung back 0.28mm again and another half hour later they were still perfectly level so I carried on.

One thing I kicked myself for not thinking of, and would probably recommend for anyone thinking about getting a bed like this, is having a slot milled away from the work area in the X and Y which you can use to indicate off to tram it all in. The dowels are a super tight to-size-need-a-dead-blow-to-get-them-in fit but I found they still left room for error and a flat edge would yield different results on the same axis depending on which row of dowels i pushed it up against. The edge of a slot takes that error away. My bed is going back on the machine soon to be finished so I'll ask him to put some slots in at the same time for whenever it all needs doing again in the future.

As it is, I think I'm close enough to square to be able to start making fixtures and have the holes match those on the bed. I'm no more than 0.1mm out over 300mm in either direction which I think will hopefully be good enough.

Surprised by the positional accuracy, assuming my indicator is accurate, every single move was on point. My 3D printer cost as much as the X6 and it doesn't move as accurately. Although I did measure 0.01mm of backlash.


Sent from my iPhone using Tapatalk

[zeeflyboy]

within 0.1mm of square across the bed is all you could ever really hope for on this machine in my experience given the gantry straightness so I'd be happy with that.

Positional accuracy is fairly good but mine was out on the steps per mm - calibration yields something like 804 steps per mm rather than the 800 it should be (the leadshines I'm using are 4000 steps per rotation). Worth checking that if you haven't already

[Mmpie]

I think you can initially get it closer. In reference to my flat edge the Y was within 0.03 and the X 0.02 after bolt tightening. But when checking that against the holes themselves it was a bit further out. How long it stays like that is what I'm wondering.

I've never properly checked motor calibration, it's probably the next thing to do properly, but I have tested small movements now and then, quite a lot yesterday while trying to square it, and always found it to move exactly as it should. Also work was always bang to size in the Y whenever I used proper finishing passes, always out slightly in X. I assumed it needed calibrating but it could just have been the massive unevenness. I never actually measured the error before I starting correcting it but it must have been somewhere close to 3mm. Starting over the centre of a hole at one end of the X left me pretty much over the edge at the other side.


Sent from my iPhone using Tapatalk

[Mmpie]

Do your electric probes automatically compensate the coordinate system alignment or is it a chore to get that set up for different jobs?


Sent from my iPhone using Tapatalk

[dharmic]

They signal when they touch something. Anything more is up to programming or how you drive the MDI.

[Mmpie]

I get that, I guess I worded my question wrong. Maybe what I should have said is

Have you been using your probes a lot to set up new jobs and is it quicker/easier to get everything keyed in and locate/align each job?

I am thinking that 9 times out of 10 you would just use the code to move the probe to find the centre/edge, rather than jogging it manually, and still do any maths yourself?

I'm wondering specifically for aligning a workpiece after a flip. I like the look of the Haimer style analog probes but when the cheapest I can find is 2x the cost of electric ones... not so much


Sent from my iPhone using Tapatalk

[RCaffin]

I have electronic probes, wobbler probes, and others. By and large, I use a ground rod in the chuck and a feeler gauge.
For flipping a job I rely on locating pins.

Cheers

[dharmic]

At the moment I hardly use it at all, if I'm honest. I just got around obstacle 1 (needing to swap plugs and pin config polarity between it and the tool height probe). I still need to do a bunch more testing and work out the probe deflection offset better.

Mostly I touch off with whatever roughing tool I'm going to use (6mm two flute) until it catches the tip when I spin it and that's good enough.

Of some interest is Fusion 360's probing tool. Unfortunately the Mach3 post needs to be extended to cope with probing but, when I get that done (or extend the UCCNC if I go that way first) it means there's an operation you can define in 360 which touches off and sets up a work offset automatically. Nice.

[RCaffin]

until it catches the tip when I spin it and that's good enough.
To be sure, that can be just as precise.
Where it fails for me is when I want to rezero to a base plate each day, for production use. I don't want to keep abrading, ever slightly, the reference edges. But that is for my needs.

Cheers
Roger

[Mmpie]

I didn't know fusion could set up a probing routine. That sounds good, anything that reduces the potential for human error is a good thing IMO, particularly with how long these machines take to do anything of note out of aluminium. Like scrapping 2 or 3 hours work because you rotated the coordinates 0.1 degrees the wrong way

Talking of probes my z probe seems to be stuck on digitise whenever the crocodile clamp is clipped to anything, whether on or off the machine. I find that the probe is quite variable though even with a double touch routine and have been using feeler gauges to set z for awhile when z height matters much, it doesn't really take much longer and is far more on point. Noticeable when using different tools on the same surface.

For x and y zero, since I got this new bed, I've chucked up my DTI and zeroed it in one of the reamed holes. Really quick and easy to keep on top of and anything not in the vice can be referenced from that point.


Sent from my iPhone using Tapatalk

[Mmpie]

Just to revisit the air compressor discussion from a couple of weeks back. Recently replaced my cheapo Chinese thing with this similar rated belt driven unit https://www.airsupplies.co.uk/abac-a...-litre-7-9-cfm

From a meter distance, noise with my box door open is 77db, with the door shut its 58db, you can hardly hear it. It's so much quieter than the cheapie and with strategic placement you could easily get away with having it in the room without an enclosure without having to worry about your ears. Operating temperatures also seem to be far lower as the temp in my box doesn't rise anywhere near as quick.

In use it happily handles the misters without having to cycle too regularly, I tend to only use one 95% of the time now, but because it's so quiet it hardly matters if it does come on, and dual mist still falls just under the allowed 50/50 cycle ratio. Also rated to use in ambient temps up to 40•C, considering the hottest my enclosure ever got in the middle of 'summer' (if such a thing exists in Britain) was 34•C under a pretty constant dual mist hammering, this thing is more than capable of keeping up.

This is the single phase 13A version, there's a more powerful single phase 16A model available too.

Just thought I'd share in case it's of any help to anyone planning on getting a new one.

Edit: mine isn't actually the one linked. It's the pro model, which is almost the same but has dual outlets and another couple of extras. For any UK based people it doesn't seem to be on their website but they have them in store and can order one over the phone.

Sent from my iPhone using Tapatalk

[bl775]

So we set up our Compressor system. The compressor seems to have enough power if it's pre-pressurised to around 120psi. Then we can have constant 40psi outflow for around 10-15 minutes. Made some cuts with this system and it has been working pretty well. This is the compressor: Air Compressors & Accessories | Portable Air Compressors | California Air Tools 5510 Ultra Quiet & Oil-Free 1 HP, 5.5 Gal. Steel Tank Air Compressor | B1898593 - GlobalIndustrial.com

[Mmpie]

Is anyone on here near their machine at the minute? Or does anyone know how much clearance there is between the bottom of the Z and the bed? As I believe it sits a little lower than the bottom of the gantry.


Sent from my iPhone using Tapatalk